Commit
•
b66819a
1
Parent(s):
3b87982
End of training
Browse files- all_results.json +13 -0
- test_results.json +13 -0
- trainer_state.json +693 -0
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.08695652173913,
|
| 3 |
+
"eval_accuracy": 0.6938775510204082,
|
| 4 |
+
"eval_accuracy_hold": 0.0,
|
| 5 |
+
"eval_accuracy_preparation": 0.8961038961038961,
|
| 6 |
+
"eval_accuracy_recovery": 0.0,
|
| 7 |
+
"eval_accuracy_stroke": 0.0,
|
| 8 |
+
"eval_accuracy_unknown": 0.9428571428571428,
|
| 9 |
+
"eval_loss": 0.929679811000824,
|
| 10 |
+
"eval_runtime": 15.1756,
|
| 11 |
+
"eval_samples_per_second": 9.687,
|
| 12 |
+
"eval_steps_per_second": 1.252
|
| 13 |
+
}
|
test_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.08695652173913,
|
| 3 |
+
"eval_accuracy": 0.6938775510204082,
|
| 4 |
+
"eval_accuracy_hold": 0.0,
|
| 5 |
+
"eval_accuracy_preparation": 0.8961038961038961,
|
| 6 |
+
"eval_accuracy_recovery": 0.0,
|
| 7 |
+
"eval_accuracy_stroke": 0.0,
|
| 8 |
+
"eval_accuracy_unknown": 0.9428571428571428,
|
| 9 |
+
"eval_loss": 0.929679811000824,
|
| 10 |
+
"eval_runtime": 15.1756,
|
| 11 |
+
"eval_samples_per_second": 9.687,
|
| 12 |
+
"eval_steps_per_second": 1.252
|
| 13 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,693 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.7238805970149254,
|
| 3 |
+
"best_model_checkpoint": "videomae-base-finetuned-good-gesturePhaseV3/checkpoint-690",
|
| 4 |
+
"epoch": 9.08695652173913,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 690,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.014492753623188406,
|
| 13 |
+
"grad_norm": 8.630314826965332,
|
| 14 |
+
"learning_rate": 1.4492753623188408e-06,
|
| 15 |
+
"loss": 1.51,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.028985507246376812,
|
| 20 |
+
"grad_norm": 7.7164106369018555,
|
| 21 |
+
"learning_rate": 2.8985507246376816e-06,
|
| 22 |
+
"loss": 1.4827,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.043478260869565216,
|
| 27 |
+
"grad_norm": 8.436184883117676,
|
| 28 |
+
"learning_rate": 4.347826086956522e-06,
|
| 29 |
+
"loss": 1.3066,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.057971014492753624,
|
| 34 |
+
"grad_norm": 5.019124507904053,
|
| 35 |
+
"learning_rate": 5.797101449275363e-06,
|
| 36 |
+
"loss": 1.3189,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.07246376811594203,
|
| 41 |
+
"grad_norm": 4.321361541748047,
|
| 42 |
+
"learning_rate": 7.246376811594203e-06,
|
| 43 |
+
"loss": 1.2353,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.08695652173913043,
|
| 48 |
+
"grad_norm": 5.294806480407715,
|
| 49 |
+
"learning_rate": 8.695652173913044e-06,
|
| 50 |
+
"loss": 1.0792,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.10144927536231885,
|
| 55 |
+
"grad_norm": 6.665987491607666,
|
| 56 |
+
"learning_rate": 9.98389694041868e-06,
|
| 57 |
+
"loss": 1.0859,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.10144927536231885,
|
| 62 |
+
"eval_accuracy": 0.5597014925373134,
|
| 63 |
+
"eval_accuracy_hold": 0.0,
|
| 64 |
+
"eval_accuracy_preparation": 1.0,
|
| 65 |
+
"eval_accuracy_recovery": 0.0,
|
| 66 |
+
"eval_accuracy_stroke": 0.0,
|
| 67 |
+
"eval_accuracy_unknown": 0.0,
|
| 68 |
+
"eval_loss": 1.2821296453475952,
|
| 69 |
+
"eval_runtime": 12.6272,
|
| 70 |
+
"eval_samples_per_second": 10.612,
|
| 71 |
+
"eval_steps_per_second": 1.346,
|
| 72 |
+
"step": 70
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.0144927536231885,
|
| 76 |
+
"grad_norm": 7.540314197540283,
|
| 77 |
+
"learning_rate": 9.822866344605476e-06,
|
| 78 |
+
"loss": 1.1573,
|
| 79 |
+
"step": 80
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 1.0289855072463767,
|
| 83 |
+
"grad_norm": 6.748633861541748,
|
| 84 |
+
"learning_rate": 9.66183574879227e-06,
|
| 85 |
+
"loss": 1.1544,
|
| 86 |
+
"step": 90
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.0434782608695652,
|
| 90 |
+
"grad_norm": 5.429695129394531,
|
| 91 |
+
"learning_rate": 9.500805152979067e-06,
|
| 92 |
+
"loss": 1.1273,
|
| 93 |
+
"step": 100
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.0579710144927537,
|
| 97 |
+
"grad_norm": 4.697651386260986,
|
| 98 |
+
"learning_rate": 9.339774557165862e-06,
|
| 99 |
+
"loss": 1.0504,
|
| 100 |
+
"step": 110
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 1.0724637681159421,
|
| 104 |
+
"grad_norm": 5.215806484222412,
|
| 105 |
+
"learning_rate": 9.178743961352658e-06,
|
| 106 |
+
"loss": 1.3456,
|
| 107 |
+
"step": 120
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 1.0869565217391304,
|
| 111 |
+
"grad_norm": 4.073849201202393,
|
| 112 |
+
"learning_rate": 9.017713365539453e-06,
|
| 113 |
+
"loss": 1.0036,
|
| 114 |
+
"step": 130
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 1.1014492753623188,
|
| 118 |
+
"grad_norm": 6.67258882522583,
|
| 119 |
+
"learning_rate": 8.85668276972625e-06,
|
| 120 |
+
"loss": 1.1816,
|
| 121 |
+
"step": 140
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 1.1014492753623188,
|
| 125 |
+
"eval_accuracy": 0.5597014925373134,
|
| 126 |
+
"eval_accuracy_hold": 0.0,
|
| 127 |
+
"eval_accuracy_preparation": 1.0,
|
| 128 |
+
"eval_accuracy_recovery": 0.0,
|
| 129 |
+
"eval_accuracy_stroke": 0.0,
|
| 130 |
+
"eval_accuracy_unknown": 0.0,
|
| 131 |
+
"eval_loss": 1.2568527460098267,
|
| 132 |
+
"eval_runtime": 10.4527,
|
| 133 |
+
"eval_samples_per_second": 12.82,
|
| 134 |
+
"eval_steps_per_second": 1.626,
|
| 135 |
+
"step": 140
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.0144927536231885,
|
| 139 |
+
"grad_norm": 6.217495918273926,
|
| 140 |
+
"learning_rate": 8.695652173913044e-06,
|
| 141 |
+
"loss": 0.8822,
|
| 142 |
+
"step": 150
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.028985507246377,
|
| 146 |
+
"grad_norm": 7.204903602600098,
|
| 147 |
+
"learning_rate": 8.53462157809984e-06,
|
| 148 |
+
"loss": 1.3084,
|
| 149 |
+
"step": 160
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 2.0434782608695654,
|
| 153 |
+
"grad_norm": 6.101042747497559,
|
| 154 |
+
"learning_rate": 8.373590982286636e-06,
|
| 155 |
+
"loss": 1.2082,
|
| 156 |
+
"step": 170
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 2.0579710144927534,
|
| 160 |
+
"grad_norm": 7.498929977416992,
|
| 161 |
+
"learning_rate": 8.212560386473431e-06,
|
| 162 |
+
"loss": 1.0949,
|
| 163 |
+
"step": 180
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 2.072463768115942,
|
| 167 |
+
"grad_norm": 5.478421688079834,
|
| 168 |
+
"learning_rate": 8.051529790660226e-06,
|
| 169 |
+
"loss": 1.1815,
|
| 170 |
+
"step": 190
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 2.0869565217391304,
|
| 174 |
+
"grad_norm": 5.867790222167969,
|
| 175 |
+
"learning_rate": 7.890499194847021e-06,
|
| 176 |
+
"loss": 1.1617,
|
| 177 |
+
"step": 200
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 2.101449275362319,
|
| 181 |
+
"grad_norm": 6.117257595062256,
|
| 182 |
+
"learning_rate": 7.729468599033817e-06,
|
| 183 |
+
"loss": 0.9776,
|
| 184 |
+
"step": 210
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 2.101449275362319,
|
| 188 |
+
"eval_accuracy": 0.5597014925373134,
|
| 189 |
+
"eval_accuracy_hold": 0.0,
|
| 190 |
+
"eval_accuracy_preparation": 1.0,
|
| 191 |
+
"eval_accuracy_recovery": 0.0,
|
| 192 |
+
"eval_accuracy_stroke": 0.0,
|
| 193 |
+
"eval_accuracy_unknown": 0.0,
|
| 194 |
+
"eval_loss": 1.183463454246521,
|
| 195 |
+
"eval_runtime": 14.5304,
|
| 196 |
+
"eval_samples_per_second": 9.222,
|
| 197 |
+
"eval_steps_per_second": 1.17,
|
| 198 |
+
"step": 210
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 3.0144927536231885,
|
| 202 |
+
"grad_norm": 4.579581260681152,
|
| 203 |
+
"learning_rate": 7.568438003220613e-06,
|
| 204 |
+
"loss": 1.0491,
|
| 205 |
+
"step": 220
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 3.028985507246377,
|
| 209 |
+
"grad_norm": 7.75211238861084,
|
| 210 |
+
"learning_rate": 7.4074074074074075e-06,
|
| 211 |
+
"loss": 1.1185,
|
| 212 |
+
"step": 230
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 3.0434782608695654,
|
| 216 |
+
"grad_norm": 5.795550346374512,
|
| 217 |
+
"learning_rate": 7.246376811594203e-06,
|
| 218 |
+
"loss": 0.8838,
|
| 219 |
+
"step": 240
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 3.0579710144927534,
|
| 223 |
+
"grad_norm": 19.677343368530273,
|
| 224 |
+
"learning_rate": 7.085346215780999e-06,
|
| 225 |
+
"loss": 1.1633,
|
| 226 |
+
"step": 250
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 3.072463768115942,
|
| 230 |
+
"grad_norm": 7.7631025314331055,
|
| 231 |
+
"learning_rate": 6.924315619967794e-06,
|
| 232 |
+
"loss": 1.1222,
|
| 233 |
+
"step": 260
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 3.0869565217391304,
|
| 237 |
+
"grad_norm": 10.160636901855469,
|
| 238 |
+
"learning_rate": 6.76328502415459e-06,
|
| 239 |
+
"loss": 1.1524,
|
| 240 |
+
"step": 270
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 3.101449275362319,
|
| 244 |
+
"grad_norm": 8.012062072753906,
|
| 245 |
+
"learning_rate": 6.602254428341386e-06,
|
| 246 |
+
"loss": 0.9867,
|
| 247 |
+
"step": 280
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 3.101449275362319,
|
| 251 |
+
"eval_accuracy": 0.6119402985074627,
|
| 252 |
+
"eval_accuracy_hold": 0.0,
|
| 253 |
+
"eval_accuracy_preparation": 0.96,
|
| 254 |
+
"eval_accuracy_recovery": 0.0,
|
| 255 |
+
"eval_accuracy_stroke": 0.0,
|
| 256 |
+
"eval_accuracy_unknown": 0.38461538461538464,
|
| 257 |
+
"eval_loss": 1.0721046924591064,
|
| 258 |
+
"eval_runtime": 10.6287,
|
| 259 |
+
"eval_samples_per_second": 12.607,
|
| 260 |
+
"eval_steps_per_second": 1.599,
|
| 261 |
+
"step": 280
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 4.0144927536231885,
|
| 265 |
+
"grad_norm": 7.033966064453125,
|
| 266 |
+
"learning_rate": 6.44122383252818e-06,
|
| 267 |
+
"loss": 1.1485,
|
| 268 |
+
"step": 290
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 4.028985507246377,
|
| 272 |
+
"grad_norm": 7.627530097961426,
|
| 273 |
+
"learning_rate": 6.280193236714976e-06,
|
| 274 |
+
"loss": 0.8141,
|
| 275 |
+
"step": 300
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 4.043478260869565,
|
| 279 |
+
"grad_norm": 27.910728454589844,
|
| 280 |
+
"learning_rate": 6.119162640901772e-06,
|
| 281 |
+
"loss": 0.8858,
|
| 282 |
+
"step": 310
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 4.057971014492754,
|
| 286 |
+
"grad_norm": 10.358675956726074,
|
| 287 |
+
"learning_rate": 5.958132045088567e-06,
|
| 288 |
+
"loss": 1.0275,
|
| 289 |
+
"step": 320
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 4.072463768115942,
|
| 293 |
+
"grad_norm": 16.66415023803711,
|
| 294 |
+
"learning_rate": 5.797101449275363e-06,
|
| 295 |
+
"loss": 1.0072,
|
| 296 |
+
"step": 330
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 4.086956521739131,
|
| 300 |
+
"grad_norm": 10.6364107131958,
|
| 301 |
+
"learning_rate": 5.6360708534621574e-06,
|
| 302 |
+
"loss": 0.9873,
|
| 303 |
+
"step": 340
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 4.101449275362318,
|
| 307 |
+
"grad_norm": 37.127933502197266,
|
| 308 |
+
"learning_rate": 5.475040257648953e-06,
|
| 309 |
+
"loss": 1.0732,
|
| 310 |
+
"step": 350
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 4.101449275362318,
|
| 314 |
+
"eval_accuracy": 0.6044776119402985,
|
| 315 |
+
"eval_accuracy_hold": 0.0,
|
| 316 |
+
"eval_accuracy_preparation": 0.7333333333333333,
|
| 317 |
+
"eval_accuracy_recovery": 0.0,
|
| 318 |
+
"eval_accuracy_stroke": 0.0,
|
| 319 |
+
"eval_accuracy_unknown": 1.0,
|
| 320 |
+
"eval_loss": 1.10855233669281,
|
| 321 |
+
"eval_runtime": 10.2262,
|
| 322 |
+
"eval_samples_per_second": 13.104,
|
| 323 |
+
"eval_steps_per_second": 1.662,
|
| 324 |
+
"step": 350
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.0144927536231885,
|
| 328 |
+
"grad_norm": 5.9557390213012695,
|
| 329 |
+
"learning_rate": 5.314009661835749e-06,
|
| 330 |
+
"loss": 1.0058,
|
| 331 |
+
"step": 360
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.028985507246377,
|
| 335 |
+
"grad_norm": 4.721350193023682,
|
| 336 |
+
"learning_rate": 5.152979066022544e-06,
|
| 337 |
+
"loss": 0.9693,
|
| 338 |
+
"step": 370
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.043478260869565,
|
| 342 |
+
"grad_norm": 10.849539756774902,
|
| 343 |
+
"learning_rate": 4.99194847020934e-06,
|
| 344 |
+
"loss": 1.0111,
|
| 345 |
+
"step": 380
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.057971014492754,
|
| 349 |
+
"grad_norm": 8.35351276397705,
|
| 350 |
+
"learning_rate": 4.830917874396135e-06,
|
| 351 |
+
"loss": 0.953,
|
| 352 |
+
"step": 390
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.072463768115942,
|
| 356 |
+
"grad_norm": 9.031652450561523,
|
| 357 |
+
"learning_rate": 4.669887278582931e-06,
|
| 358 |
+
"loss": 0.957,
|
| 359 |
+
"step": 400
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.086956521739131,
|
| 363 |
+
"grad_norm": 28.449277877807617,
|
| 364 |
+
"learning_rate": 4.508856682769726e-06,
|
| 365 |
+
"loss": 0.8448,
|
| 366 |
+
"step": 410
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 5.101449275362318,
|
| 370 |
+
"grad_norm": 8.963016510009766,
|
| 371 |
+
"learning_rate": 4.347826086956522e-06,
|
| 372 |
+
"loss": 0.9688,
|
| 373 |
+
"step": 420
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 5.101449275362318,
|
| 377 |
+
"eval_accuracy": 0.6865671641791045,
|
| 378 |
+
"eval_accuracy_hold": 0.0,
|
| 379 |
+
"eval_accuracy_preparation": 0.9333333333333333,
|
| 380 |
+
"eval_accuracy_recovery": 0.0,
|
| 381 |
+
"eval_accuracy_stroke": 0.0,
|
| 382 |
+
"eval_accuracy_unknown": 0.8461538461538461,
|
| 383 |
+
"eval_loss": 0.9288297295570374,
|
| 384 |
+
"eval_runtime": 10.1881,
|
| 385 |
+
"eval_samples_per_second": 13.153,
|
| 386 |
+
"eval_steps_per_second": 1.669,
|
| 387 |
+
"step": 420
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 6.0144927536231885,
|
| 391 |
+
"grad_norm": 6.095980167388916,
|
| 392 |
+
"learning_rate": 4.186795491143318e-06,
|
| 393 |
+
"loss": 0.7678,
|
| 394 |
+
"step": 430
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 6.028985507246377,
|
| 398 |
+
"grad_norm": 6.365720272064209,
|
| 399 |
+
"learning_rate": 4.025764895330113e-06,
|
| 400 |
+
"loss": 0.7864,
|
| 401 |
+
"step": 440
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 6.043478260869565,
|
| 405 |
+
"grad_norm": 9.036483764648438,
|
| 406 |
+
"learning_rate": 3.864734299516908e-06,
|
| 407 |
+
"loss": 0.9962,
|
| 408 |
+
"step": 450
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 6.057971014492754,
|
| 412 |
+
"grad_norm": 7.131446361541748,
|
| 413 |
+
"learning_rate": 3.7037037037037037e-06,
|
| 414 |
+
"loss": 0.8646,
|
| 415 |
+
"step": 460
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 6.072463768115942,
|
| 419 |
+
"grad_norm": 20.672754287719727,
|
| 420 |
+
"learning_rate": 3.5426731078904997e-06,
|
| 421 |
+
"loss": 1.0675,
|
| 422 |
+
"step": 470
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 6.086956521739131,
|
| 426 |
+
"grad_norm": 7.654391765594482,
|
| 427 |
+
"learning_rate": 3.381642512077295e-06,
|
| 428 |
+
"loss": 0.8623,
|
| 429 |
+
"step": 480
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 6.101449275362318,
|
| 433 |
+
"grad_norm": 5.464959144592285,
|
| 434 |
+
"learning_rate": 3.22061191626409e-06,
|
| 435 |
+
"loss": 0.7733,
|
| 436 |
+
"step": 490
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 6.101449275362318,
|
| 440 |
+
"eval_accuracy": 0.6567164179104478,
|
| 441 |
+
"eval_accuracy_hold": 0.0,
|
| 442 |
+
"eval_accuracy_preparation": 0.8266666666666667,
|
| 443 |
+
"eval_accuracy_recovery": 0.0,
|
| 444 |
+
"eval_accuracy_stroke": 0.0,
|
| 445 |
+
"eval_accuracy_unknown": 1.0,
|
| 446 |
+
"eval_loss": 0.9749326705932617,
|
| 447 |
+
"eval_runtime": 10.4474,
|
| 448 |
+
"eval_samples_per_second": 12.826,
|
| 449 |
+
"eval_steps_per_second": 1.627,
|
| 450 |
+
"step": 490
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 7.0144927536231885,
|
| 454 |
+
"grad_norm": 7.121547698974609,
|
| 455 |
+
"learning_rate": 3.059581320450886e-06,
|
| 456 |
+
"loss": 0.78,
|
| 457 |
+
"step": 500
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 7.028985507246377,
|
| 461 |
+
"grad_norm": 6.011748313903809,
|
| 462 |
+
"learning_rate": 2.8985507246376816e-06,
|
| 463 |
+
"loss": 1.0792,
|
| 464 |
+
"step": 510
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 7.043478260869565,
|
| 468 |
+
"grad_norm": 7.9175567626953125,
|
| 469 |
+
"learning_rate": 2.7375201288244767e-06,
|
| 470 |
+
"loss": 0.8386,
|
| 471 |
+
"step": 520
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 7.057971014492754,
|
| 475 |
+
"grad_norm": 7.427879333496094,
|
| 476 |
+
"learning_rate": 2.576489533011272e-06,
|
| 477 |
+
"loss": 0.6215,
|
| 478 |
+
"step": 530
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 7.072463768115942,
|
| 482 |
+
"grad_norm": 4.626201629638672,
|
| 483 |
+
"learning_rate": 2.4154589371980677e-06,
|
| 484 |
+
"loss": 0.8965,
|
| 485 |
+
"step": 540
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 7.086956521739131,
|
| 489 |
+
"grad_norm": 14.431778907775879,
|
| 490 |
+
"learning_rate": 2.254428341384863e-06,
|
| 491 |
+
"loss": 0.9818,
|
| 492 |
+
"step": 550
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 7.101449275362318,
|
| 496 |
+
"grad_norm": 32.45638656616211,
|
| 497 |
+
"learning_rate": 2.093397745571659e-06,
|
| 498 |
+
"loss": 0.8798,
|
| 499 |
+
"step": 560
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 7.101449275362318,
|
| 503 |
+
"eval_accuracy": 0.6865671641791045,
|
| 504 |
+
"eval_accuracy_hold": 0.0,
|
| 505 |
+
"eval_accuracy_preparation": 0.92,
|
| 506 |
+
"eval_accuracy_recovery": 0.0,
|
| 507 |
+
"eval_accuracy_stroke": 0.0,
|
| 508 |
+
"eval_accuracy_unknown": 0.8846153846153846,
|
| 509 |
+
"eval_loss": 0.8742198944091797,
|
| 510 |
+
"eval_runtime": 10.8069,
|
| 511 |
+
"eval_samples_per_second": 12.399,
|
| 512 |
+
"eval_steps_per_second": 1.573,
|
| 513 |
+
"step": 560
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 8.014492753623188,
|
| 517 |
+
"grad_norm": 7.238797664642334,
|
| 518 |
+
"learning_rate": 1.932367149758454e-06,
|
| 519 |
+
"loss": 0.9141,
|
| 520 |
+
"step": 570
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 8.028985507246377,
|
| 524 |
+
"grad_norm": 5.496193885803223,
|
| 525 |
+
"learning_rate": 1.7713365539452498e-06,
|
| 526 |
+
"loss": 0.7796,
|
| 527 |
+
"step": 580
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 8.043478260869565,
|
| 531 |
+
"grad_norm": 7.790290355682373,
|
| 532 |
+
"learning_rate": 1.610305958132045e-06,
|
| 533 |
+
"loss": 0.9178,
|
| 534 |
+
"step": 590
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 8.057971014492754,
|
| 538 |
+
"grad_norm": 10.017251014709473,
|
| 539 |
+
"learning_rate": 1.4492753623188408e-06,
|
| 540 |
+
"loss": 0.8282,
|
| 541 |
+
"step": 600
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 8.072463768115941,
|
| 545 |
+
"grad_norm": 11.363780975341797,
|
| 546 |
+
"learning_rate": 1.288244766505636e-06,
|
| 547 |
+
"loss": 0.8729,
|
| 548 |
+
"step": 610
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 8.08695652173913,
|
| 552 |
+
"grad_norm": 5.4790778160095215,
|
| 553 |
+
"learning_rate": 1.1272141706924316e-06,
|
| 554 |
+
"loss": 0.6978,
|
| 555 |
+
"step": 620
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 8.101449275362318,
|
| 559 |
+
"grad_norm": 9.819117546081543,
|
| 560 |
+
"learning_rate": 9.66183574879227e-07,
|
| 561 |
+
"loss": 0.7771,
|
| 562 |
+
"step": 630
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 8.101449275362318,
|
| 566 |
+
"eval_accuracy": 0.7014925373134329,
|
| 567 |
+
"eval_accuracy_hold": 0.0,
|
| 568 |
+
"eval_accuracy_preparation": 0.92,
|
| 569 |
+
"eval_accuracy_recovery": 0.0,
|
| 570 |
+
"eval_accuracy_stroke": 0.0,
|
| 571 |
+
"eval_accuracy_unknown": 0.9615384615384616,
|
| 572 |
+
"eval_loss": 0.845240592956543,
|
| 573 |
+
"eval_runtime": 10.5946,
|
| 574 |
+
"eval_samples_per_second": 12.648,
|
| 575 |
+
"eval_steps_per_second": 1.605,
|
| 576 |
+
"step": 630
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 9.014492753623188,
|
| 580 |
+
"grad_norm": 17.136520385742188,
|
| 581 |
+
"learning_rate": 8.051529790660226e-07,
|
| 582 |
+
"loss": 0.8034,
|
| 583 |
+
"step": 640
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 9.028985507246377,
|
| 587 |
+
"grad_norm": 8.827512741088867,
|
| 588 |
+
"learning_rate": 6.44122383252818e-07,
|
| 589 |
+
"loss": 0.6385,
|
| 590 |
+
"step": 650
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 9.043478260869565,
|
| 594 |
+
"grad_norm": 9.634197235107422,
|
| 595 |
+
"learning_rate": 4.830917874396135e-07,
|
| 596 |
+
"loss": 0.8629,
|
| 597 |
+
"step": 660
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 9.057971014492754,
|
| 601 |
+
"grad_norm": 10.844426155090332,
|
| 602 |
+
"learning_rate": 3.22061191626409e-07,
|
| 603 |
+
"loss": 0.9232,
|
| 604 |
+
"step": 670
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 9.072463768115941,
|
| 608 |
+
"grad_norm": 11.182330131530762,
|
| 609 |
+
"learning_rate": 1.610305958132045e-07,
|
| 610 |
+
"loss": 0.6174,
|
| 611 |
+
"step": 680
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 9.08695652173913,
|
| 615 |
+
"grad_norm": 9.969412803649902,
|
| 616 |
+
"learning_rate": 0.0,
|
| 617 |
+
"loss": 1.0448,
|
| 618 |
+
"step": 690
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 9.08695652173913,
|
| 622 |
+
"eval_accuracy": 0.7238805970149254,
|
| 623 |
+
"eval_accuracy_hold": 0.0,
|
| 624 |
+
"eval_accuracy_preparation": 0.96,
|
| 625 |
+
"eval_accuracy_recovery": 0.0,
|
| 626 |
+
"eval_accuracy_stroke": 0.0,
|
| 627 |
+
"eval_accuracy_unknown": 0.9615384615384616,
|
| 628 |
+
"eval_loss": 0.835747241973877,
|
| 629 |
+
"eval_runtime": 11.1151,
|
| 630 |
+
"eval_samples_per_second": 12.056,
|
| 631 |
+
"eval_steps_per_second": 1.529,
|
| 632 |
+
"step": 690
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 9.08695652173913,
|
| 636 |
+
"step": 690,
|
| 637 |
+
"total_flos": 6.856030795503698e+18,
|
| 638 |
+
"train_loss": 0.9979172112285227,
|
| 639 |
+
"train_runtime": 1188.5395,
|
| 640 |
+
"train_samples_per_second": 4.644,
|
| 641 |
+
"train_steps_per_second": 0.581
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 9.08695652173913,
|
| 645 |
+
"eval_accuracy": 0.6938775510204082,
|
| 646 |
+
"eval_accuracy_hold": 0.0,
|
| 647 |
+
"eval_accuracy_preparation": 0.8961038961038961,
|
| 648 |
+
"eval_accuracy_recovery": 0.0,
|
| 649 |
+
"eval_accuracy_stroke": 0.0,
|
| 650 |
+
"eval_accuracy_unknown": 0.9428571428571428,
|
| 651 |
+
"eval_loss": 0.929679811000824,
|
| 652 |
+
"eval_runtime": 14.0873,
|
| 653 |
+
"eval_samples_per_second": 10.435,
|
| 654 |
+
"eval_steps_per_second": 1.349,
|
| 655 |
+
"step": 690
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 9.08695652173913,
|
| 659 |
+
"eval_accuracy": 0.6938775510204082,
|
| 660 |
+
"eval_accuracy_hold": 0.0,
|
| 661 |
+
"eval_accuracy_preparation": 0.8961038961038961,
|
| 662 |
+
"eval_accuracy_recovery": 0.0,
|
| 663 |
+
"eval_accuracy_stroke": 0.0,
|
| 664 |
+
"eval_accuracy_unknown": 0.9428571428571428,
|
| 665 |
+
"eval_loss": 0.929679811000824,
|
| 666 |
+
"eval_runtime": 15.1756,
|
| 667 |
+
"eval_samples_per_second": 9.687,
|
| 668 |
+
"eval_steps_per_second": 1.252,
|
| 669 |
+
"step": 690
|
| 670 |
+
}
|
| 671 |
+
],
|
| 672 |
+
"logging_steps": 10,
|
| 673 |
+
"max_steps": 690,
|
| 674 |
+
"num_input_tokens_seen": 0,
|
| 675 |
+
"num_train_epochs": 9223372036854775807,
|
| 676 |
+
"save_steps": 500,
|
| 677 |
+
"stateful_callbacks": {
|
| 678 |
+
"TrainerControl": {
|
| 679 |
+
"args": {
|
| 680 |
+
"should_epoch_stop": false,
|
| 681 |
+
"should_evaluate": false,
|
| 682 |
+
"should_log": false,
|
| 683 |
+
"should_save": true,
|
| 684 |
+
"should_training_stop": true
|
| 685 |
+
},
|
| 686 |
+
"attributes": {}
|
| 687 |
+
}
|
| 688 |
+
},
|
| 689 |
+
"total_flos": 6.856030795503698e+18,
|
| 690 |
+
"train_batch_size": 8,
|
| 691 |
+
"trial_name": null,
|
| 692 |
+
"trial_params": null
|
| 693 |
+
}
|